Evidence from genome wide association studies implicates reduced control of Epstein-Barr virus infection in multiple sclerosis susceptibility

Ali Afrasiabi†,
Grant P. Parnell†,
Nicole Fewings,
Stephen D. Schibeci,
Monica A. Basuki,
Ramya Chandramohan,
Yuan Zhou,
Bruce Taylor,
David A. Brown,
Sanjay Swaminathan,
Fiona C. McKay,
Graeme J. Stewart and
David R. BoothEmail author View ORCID ID profile

†Contributed equally

Genome Medicine201911:26

https://doi.org/10.1186/s13073-019-0640-z

Received: 28 October 2018
Accepted: 10 April 2019
Published: 30 April 2019

Abstract

Background

Genome wide association studies have identified > 200 susceptibility loci accounting for much of the heritability of multiple sclerosis (MS). Epstein-Barr virus (EBV), a memory B cell tropic virus, has been identified as necessary but not sufficient for development of MS. The molecular and immunological basis for this has not been established. Infected B cell proliferation is driven by signalling through the EBV produced cell surface protein LMP1, a homologue of the MS risk gene CD40.

Methods

We have investigated transcriptomes of B cells and EBV-infected B cells at Latency III (LCLs) and identified MS risk genes with altered expression on infection and with expression levels associated with the MS risk genotype (LCLeQTLs). The association of LCLeQTL genomic burden with EBV phenotypes in vitro and in vivo was examined. The risk genotype effect on LCL proliferation with CD40 stimulation was assessed.

Results

These LCLeQTL MS risk SNP:gene pairs (47 identified) were over-represented in genes dysregulated between B and LCLs (p < 1.53 × 10−4), and as target loci of the EBV transcription factor EBNA2 (p < 3.17 × 10−16). Overall genetic burden of LCLeQTLs was associated with some EBV phenotypes but not others. Stimulation of the CD40 pathway by CD40L reduced LCL proliferation (p < 0.001), dependent on CD40 and TRAF3 MS risk genotypes. Both CD40 and TRAF3 risk SNPs are in binding sites for the EBV transcription factor EBNA2, with expression of each correlated with EBNA2 expression dependent on genotype.